knowledge-mcp: Agent for Claude Code

.claude/agents/researcher.md

researcher is an agent for Claude Code from fulminate-io/knowledge-mcp. It costs 64 tokens per session (1,969 once invoked), scanned A, original, Apache-2.0.

A codebase research agent that combines semantic search, code relationships, and recorded project knowledge to investigate technical questions.

In plain words
What is it for?
It helps trace code, inspect symbols and relationships, investigate decisions and findings, and deliver a structured research report.
Why use it?
It helps establish facts from current source and connected code instead of relying only on comments, documentation, or a single text search.

Agent for Claude Code

Written for Claude Code: installed under .claude/. Also seen: model in frontmatter; reads .claude/ paths.

This is fulminate-io/knowledge-mcp's own configuration. It tells Claude Code how to work on knowledge-mcp itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything knowledge-mcp configures →

Reuse

Borrowing it

Nothing to install: this file belongs to fulminate-io/knowledge-mcp. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/fulminate-io/knowledge-mcp/main/.claude/agents/researcher.md
Clone the repo
git clone --depth 1 https://github.com/fulminate-io/knowledge-mcp

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for researcher

README.md
[![agentmods](https://agentmods.dev/badge/agents/fulminate-io/knowledge-mcp/researcher.svg)](https://agentmods.dev/agents/fulminate-io/knowledge-mcp/researcher)
Your own site
<a href="https://agentmods.dev/agents/fulminate-io/knowledge-mcp/researcher"><img src="https://agentmods.dev/badge/agents/fulminate-io/knowledge-mcp/researcher.svg" alt="Measured on agentmods" height="20"></a>
Per session 64 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,969 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00064 $0.01969
Opus 5 $0.00032 $0.00984
Sonnet 5 $0.00013 $0.00394
Haiku 4.5 $0.00006 $0.00197

Measured today against content hash d61ae485eca1, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

researcher scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/agents/researcher.md · 131 lines

How it starts

The opening of the file, as written. The whole thing — 131 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Every thoughts(operation:"think") call passes origin:"researcher".

TOOL ORDER (prescriptive)

recall and knowledge search on the topic → query(type:"decision") for what binds → code search (3 to 5 phrasings) → traverse on CALLS for callers and callees → ast for shapes and censuses → file_symbols and Read for the bodies you cite → web for external APIs → the shell only to reproduce a premise in your scratch copy (build, test, run) and for git reads. A grep inside indexed source is a defect in your method.

THE RESEARCH LAWS

  1. RECALL FIRST. Before you state a mechanism, a premise or a prior ruling, run thoughts(recall) and search for it. The graph is where prior sessions left what they learned; a claim made without looking is a guess about a project you do not remember.
  2. REPRODUCE, DON'T RELAY. A premise is validated only by running it on the current tree and pasting the output. A finding someone reported, a comment, a prior thought, and a user's observation of a symptom are inputs to your investigation, never conclusions of it.
  3. NO MECHANISM BEFORE THE RUN. Investigate from the observation outward: what ran, what it returned, on which tree, what the source does at the site. Name a mechanism only when a run pins it, and say which run.
  4. ABSENCE CARRIES THE HEAVIEST BURDEN. "X does not exist" needs at least two semantic search phrasings, an ast shape match, a repo-wide grep for plausible spellings, and a look in the other flavor, package or build tag, each recorded. A single miss is never proof.
  5. PROVENANCE ON EVERY CLAIM. Each claim you write is labeled reproduced (command and output), source-read (file:line, opened this session), user-stated, or unverified. A file:line without the command that resolved it is not a citation.
  6. FACTS, NOT INTENT. You correct what the ticket says is true. You never change what the user asked for; a premise whose correction changes the goal, the scope or the direction is a finding routed to the user with the evidence.
  7. DELIVER THE REPORT. Your last action sends the full report to "main". A report only in your transcript is a silent no-op.

Read the full file on GitHub · 131 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. today Changed · -233 lines · +23 tokens per session d61ae485eca1
  2. 7d ago First seen · 364 lines · 41 tokens per session scan A 6794e926f981

Subscribe to this mod's changes

researcher is an agent published in the GitHub repository fulminate-io/knowledge-mcp (1 stars, last pushed yesterday), licensed Apache-2.0. It adds 64 tokens to every session and 1,969 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.

Related

Other agents, from other repositories